-
大佬，我在测试model时发现车辆在变道时，检测出的车道会扭曲变形严重，变道结束后，效果就恢复良好了，分析了下，原因感觉是变道过程中车辆骑在车道线上行驶，导致模型对这条骑着的车道线进行分类时，无法区分是左车道还是右车道，所以产生一个中间临界状态，不知这个解释是否正确，以及这种变道过程中检测效果变差的case该如何解决呢， 拜谢大佬！
-
I have managed to run on my own images, however, I can not understand why end lines are a little bit shifted. It seems to me that pretrained weights are working well, but there are maybe some postproc…
-
when i set the "resume" value in the config file, some key errors occurs as below:
==> Resume model from ./checkpoint/culane_18.pth
Traceback (most recent call last):
File "train.py", line 126,…
-
I have no idea on how to change tusimple label.json to CULane label.txt, will you please give me some advice ?
-
Hello, I got a 69.9 on the CULane dataset using a single GPU card with your default settings. I visualized the lane detection results and found that the model performed well in some ideal scene. But t…
-
Hi @cfzd , thanks for your great work, it's really impressive!
When I train your model on my own data, I notice that my own data doesn't have instance-level land ids like CULane or TuSimple, and it i…
-
Hi, bro.
I'm appreciated for your work on lane detection. when I run train.py, I found this error:
on$ **python train.py configs/culane. py**
Traceback (most recent call last):
File "train.py"…
-
Hi,
I've been recently experimenting with using different feature extractor, in this case MobilenetV3 and efficientNet-B2.
Results with MobilenetV3 was acceptable but in the case of efficientNet I…
-
Hi, I use the default config file(culane.py) in your repo to train the model under a single GTX 1080Ti and got an F-measure of 46.1 which is far from your model 69.7. Could you share some hyper-parame…
-
I download the code on April 18, 2020. I configured the data path as is required. Then I launch the traning process by:
`python3 train.py`
After like 2 minutes, just 100 iterations, the batch lo…